pyGPGO.surrogates.BoostedTrees module

class pyGPGO.surrogates.BoostedTrees.BoostedTrees(q1=0.16, q2=0.84, **params)

Bases: object

Gradient boosted trees as surrogate model for Bayesian Optimization. Uses quantile regression for an estimate of the ‘posterior’ variance. In practice, the std is computed as (q2 - q1) / 2. Relies on sklearn.ensemble.GradientBoostingRegressor

Parameters:

• q1 (float) – First quantile.
• q2 (float) – Second quantile
• params (tuple) – Extra parameters to pass to GradientBoostingRegressor

__init__(q1=0.16, q2=0.84, **params)

Gradient boosted trees as surrogate model for Bayesian Optimization. Uses quantile regression for an estimate of the ‘posterior’ variance. In practice, the std is computed as (q2 - q1) / 2. Relies on sklearn.ensemble.GradientBoostingRegressor

Parameters:

• q1 (float) – First quantile.
• q2 (float) – Second quantile
• params (tuple) – Extra parameters to pass to GradientBoostingRegressor

fit(X, y)

Fit a GBM model to data X and targets y.

Parameters:

• X (array-like) – Input values.
• y (array-like) – Target values.

predict(Xstar, return_std=True)

Predicts ‘posterior’ mean and variance for the GBM model.

Parameters:

• Xstar (array-like) – Input values.
• return_std (bool, optional) – Whether to return posterior variance estimates. Default is True.
• eps (float, optional) – Floating precision value for negative variance estimates. Default is 1e-6

Returns:

• array-like – Posterior predicted mean.
• array-like – Posterior predicted std

update(xnew, ynew)

Updates the internal RF model with observations xnew and targets ynew.

Parameters:

• xnew (array-like) – New observations.
• ynew (array-like) – New targets.